With so much attention on, and some hype around, generative AI, organizational leaders need to understand how they can deploy it in their workplace to add real value. If your organization hasn’t yet dipped its toe in the AI waters, focus on getting AI-ready. Use the following considerations and examples of practical applications of AI to spark inspiration and motivation.
Add business value with generative AI: A framework for identifying AI use cases
It can be useful to think about AI use cases from a perspective that considers the complexity of implementation. From least to most complex to adopt, generative AI applications include those that are:
- Productivity-related.
- OEM-embedded.
- Implemented to address specific business problems — what we call, intentional use.
While this isn’t an exhaustive list, this will give you a baseline spectrum of categories to help you focus efforts, determine where to dedicate resources, and pinpoint what makes the best sense for your organization.
What are some common productivity-related use cases of generative AI?
Generative AI can help automate or supplement routine tasks to improve efficiency and resource allocation and optimize business processes. For organizations facing budget pressure, productivity-related AI applications can help maintain and improve output, service delivery, and client experience without substantially increasing costs.
A baseline use case could be asking an AI chatbot like ChatGPT or Microsoft Copilot to analyze one or more documents or spreadsheets for key insights. For example, construction and other industries are leveraging generative AI to analyze contracts and identify federal components subject to the new cybersecurity maturity model certification (CMMC) framework.
Community colleges and other higher education institutions can use generative AI for fraud detection and prevention, including to identify so-called “ghost students” and enrollment and financial aid fraud. More broadly, generative AI can assist with other types of anomaly detection that could indicate fraud, such as in expense reimbursement requests or staff credit card charges, helping to mitigate risk of financial loss.
What are some common use cases for AI that are OEM-embedded?
This category of generative AI use case covers AI that OEMs are increasingly integrating into their products, such as Salesforce, Workday, Oracle, and ServiceNow. This type of embedded AI can provide for rapid time-to-deployment, often seamlessly integrating existing business processes. In many smart city and IoT (Internet of Things) platforms, AI is operating under the hood.
Smart procurement and automated invoice processing are two specific, real-world AI use cases. With AI support, systems can generate purchase orders and process vendor invoices. Your accounts payable clerks then can be the “human in the loop” who review for accuracy and advance the process to authorization or final approval. AI apps also can analyze procurement data to identify cost-saving opportunities, optimize supplier selection, and ensure compliance with procurement policies.
Autonomous drones coupled with AI-enabled aerial intelligence systems are being used for multiple construction site applications. Use cases range from project inventory oversight and obstacle avoidance to safety metrics monitoring.
Project scheduling and resource planning is a significantly time-consuming activity for construction businesses, and AI-enabled platforms are also being used to speed time to insight as project timelines shift. AI solutions can assist with construction project simulation, optimization, and resource planning, including balancing skilled and unskilled labor, materials and equipment, and project timelines.
What are some purpose-built AI use cases to address specific business problems?
This category has the least uptake to date and is the most complex to implement but also potentially can provide a competitive advantage and be a catalyst for your organization’s digital transformation.
Some large metropolitan areas are implementing smart traffic management, leveraging AI to process and analyze data from traffic cameras and other sources. The data is then used to adjust traffic signal timing as real-time conditions change to reduce congestion.
Public safety use cases of generative AI include analyzing crowdsourced traffic, weather, and crime information from mobile phones, doorbell apps, municipal cameras, and real-time mapping and traffic data sources. Generative AI analysis and predictive capabilities can help municipalities address crime patterns, optimize the number and location of law enforcement units on the street, or better plan shift schedules. Facial and license plate recognition can expedite police reports and other administrative tasks.
Construction firms are beginning to adopt generative AI in project bidding and cost estimating, drawing on past project specs and margins to inform their competitive bidding strategy. Construction businesses have plenty of historic job-specific data — the trick is to be able to mine it and leverage the insights. This can reduce staff time on creating bids, although we humans aren’t removed from the process entirely. It’s critical to have an experienced team member validate the suggestions AI-enabled tools generate — AI is not the easy button.
How do you identify generative AI use cases?
As you begin to consider your AI use cases, start by thinking about the business issues you want to address rather than the technology. In other words, identify the problem first, then back into the technology.
Once you have your list of business issues, consider whether AI can in fact address each one. Some business problems can be solved with AI, and some can’t. Then you can begin to shape your use cases and determine the best tools.
Leverage the requirements and innovation cycle you likely already use for adopting any other technology. The good news is, depending on the use case, you won’t necessarily have to invest in something shiny and new; a platform you already license may have AI-enabled features you haven’t yet investigated or there may already be a pre-built solution for your use case.
Is your data AI ready?
Before considering where and how AI could benefit your business, it’s crucial to ensure your data is “AI-ready.” Is your data free from errors, inconsistencies, and biases? Is your organization collecting the right data for the AI applications you’re thinking about, and is that data available and complete? Preparing data for any AI application is the first step.
